Giving life to GPRA

It's a time-tested truism among bureaucrats, John Mercer knows — if you don't like a particular policy, wait awhile. Presidents and Cabinet members come and go, members of Congress lose re-election or retire, circumstances change. Only the bureaucracy endures.

Thus, the odds seemed stacked against GPRA, the 1993 Government Performance and Results Act, which Mercer wrote in 1990 as Republican counsel to the Senate Governmental Affairs Committee.

Passed amid the "re.inventing government" fervor of the Clinton administration, GPRA aimed to impose "performance-based budgeting" on federal agencies. Rather than simply show that the appropriate number of dollars had been spent on the right programs, agencies would have to show how spending money was producing results.

GPRA represented a radical change for government budget officials, who, Mercer said, are accustomed to justifying budgets by touting their programs' good intentions, not their performance. Presented with the idea of tying spending to performance, agency budgeters balked.

A congressional audit of 24 major agencies in 1997 showed just how poorly agencies were complying with GPRA. Two agencies earned a grade of C, three merited a D and the other 19 got an F. But bad grades prompted no obvious reform. Two years later, the average agency grade was still failing.

"There was a lot of sentiment within the agencies that this would go away with the change of administration," Mercer said.

What GPRA antagonists failed to realize, however, was that "this was really a Republican initiative," Mercer said with a smile. President Bush may be even more gung-ho for GPRA than was his Democratic predecessor.

That's great news for Mercer, the self-proclaimed "father of GPRA."

He left the Senate staff in 1997 to work as Federal Data Corp.'s deputy director for government performance, then launched his own consulting business in July. His specialty: teaching government agencies how to implement performance-based budgeting, the philosophical heart of GPRA.

Opportunities abound, Mercer said, because GPRA did not go away — and government managers still don't understand it.

He drives a sport utility vehicle with license plates that read "GPRA."

Mercer refined the idea of performance-based budgeting during the 1980s, while he was mayor of Sunnyvale, Calif., and he pulled out Sunnyvale's police budget to demonstrate how performance-based budgeting works.

The police budget is unusual in a number of respects, one being its degree of detail. More than 80 line items disclose to the penny how much the city plans to spend on items ranging from salaries and overtime to ammunition and office supplies.

Armed with dozens of such details, Mercer explained, it becomes possible to calculate the cost of the police department's "outputs," such as the cost per hour of patrolling city streets, transporting prisoners or providing crime- prevention services.

Knowing the unit cost of outputs makes it possible to measure how budget dollars spent translate into results achieved, Mercer said. And knowing the dollar costs of results makes it possible to predict with a degree of accuracy how increases or decreases in spending in various parts of the budget will lead to different results.

But the many line items and to-the-penny details — "granularity," Mercer calls it — are only part of performance-based budgeting. Sunnyvale's 17-page police budget begins with a "program outcome statement," almost a declaration of the police department's purpose: "Provide a safe community and a feeling of security among the citizens" by protecting lives, property and rights, preventing crime, ensuring traffic safety and so on "to the satisfaction of citizens."

The budget sets "program outcome measures" for determining whether the department is meeting its goals. The 1999 budget, for example, included measures such as keeping the local crime rate 54 percent below the national crime rate and keeping response times to urgent police calls under seven minutes 90 percent of the time.

There is also a service delivery plan that shows how budget dollars are allocated among dozens of activities, from collecting and processing physical evidence of crimes to providing crime- prevention services.

From its wealth of small details to its sweeping goal of providing a safe community, the police budget provides great "transparency" into how the police department spends its money and what police spending is expected to achieve, Mercer said.

In some instances, performance budgeting may be easier for local entities than for federal agencies, Mercer conceded. In law enforcement, for example, federal agencies turn over billions of dollars each year to state and local agencies, but federal agencies often have little or no control over how their money is ultimately spent.

That makes it hard for federal agencies to show how their spending contributed to crime reduction — or why, if it did not, it wasn't their fault, Mercer said.

An objection agency officials often raise against performance-based budgeting is that their budgets could be cut if programs they do not control fail to produce results.

"We ought not to be too rigid about holding their feet to the fire for not meeting goals — at least initially," Mercer said. Agencies can still use performance-based budgeting to meas.ure their performance as distributors of federal largess. In that case, outcomes would not be crimes prevented, but such measures as whether the proper disbursements were made to the right localities on time, and whether ap.propriate audits and oversight were performed.

But ultimately, performance-based budgeting is intended to weed out programs that do not achieve desired results.

Agencies hoping GPRA would go away cannot have been heartened by the Bush administration's rhetoric so far. Top officials in the Office of Management and Budget have repeatedly identified performance-based budgeting as a top priority, and Bush has said that "nonperforming activities should be reformed or terminated."

Last winter, Mercer explained performance-based budgeting in a letter to OMB Director Mitchell Daniels Jr., an acquaintance from Mercer's days as Sunnyvale mayor, when Daniels was President Reagan's liaison to state and local governments.

In May, GPRA was the subject of a Cabinet meeting, and in June, Mercer was invited to present senior managers at OMB with a GPRA tutorial.

"The administration is serious about integrating performance with budgets," but so far OMB has not provided agencies with much guidance about how to do it, Mercer said.

"The best thing to do is be a leader," he said. The agencies that develop successful performance-based budgeting "will be looked upon favorably," and those that hesitate will be forced to adopt practices developed by those that get out in front. "There is a first-mover advantage," he said.

Mercer is counting on it. Recently, he has been advising the Bureau of Alcohol, Tobacco and Firearms on how to adopt performance-based budgeting and has fielded inquiries from the Coast Guard, the Environmental Protection Agency and a third "major independent agency" he declined to identify.

***

The John Mercer File

Position: Self-employed consultant.

Previous job: Federal Data Corp. deputy director for government performance. Other resume highlights: Republican counsel, Senate Governmental Affairs Committee, 1989 to 1997; Department of Housing and Urban Development deputy assistant secretary for program policy development and evaluation, 1988 to 1989; city council member and mayor of Sunnyvale, Calif., 1979 to 1987.

Education: Bachelor's degree in public administration from San Jose State University and law degree from University of Notre Dame.

Quote: "The best thing to do is be a leader."